基于模糊PID控制的网络异常流量监控设计  

Network Abnormal Traffic Monitoring Based on Fuzzy PID Control

作　　者：王俊士[1] 

机构地区：[1]河南大学濮阳工学院数学与信息工程系,河南濮阳457000

出　　处：《科技通报》2016年第5期142-146,共5页Bulletin of Science and Technology

摘　　要：在大型网络系统中,会因为网络拥堵或者病毒入侵产生网络异常流量,通过对网络异常流量的监控系统设计保障网络的安全稳定。传统的网络异常流量监控系统采用BP神经网络控制算法,由于BP神经网络的结构单一性,对异常流量中干扰性较强的数据难以有效检测发现,提出一种基于改进的模糊PID控制的网络流量异常监控系统设计方法,在流量传输模型和时间序列分析模型构建的基础上,提取异常流量序列的定量递归熵特征,采用模糊PID神经网络控制方法进行流量监控系统设计,把监控到的异常流量通过链路信道回馈给输出层。仿真结果表明,采用该方法进行网络异常流量监控,能准确检测出异常的频谱特征,从而在时域上进行追溯定位,实现流量异常点的准确检测,对异常流量的准确检测概率比传统方法高。In the large network system, because of the network congestion or the virus invasion to produceabnormal network traffic, network traffic monitoring system design and security network security andstability. BP neural network control algorithm is used in the traditional network anomaly trafficmonitoring system. The BP neural network＇s structure is simple, and it is difficult to detect the abnormaltraffic. Based on the improved fuzzy PID control system, the traffic flow model and time series analysismodel is built. The quantitative recursive entropy feature is extracted. The PID neural network controlmethod is used to design the traffic monitoring system. The simulation results show that this method candetect the abnormal spectrum characteristics accurately, and can be used to detect the abnormal trafficin time domain.

关 键 词：模糊PID控制 网络 异常流量 监控系统 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]